A Reasonable Sleep

Sleeping with babies could ward off sudden infant death syndrome.

By Meredith F. Small|Wednesday, April 01, 1992
Three-month-old Jenny lies in the crook of her mother’s arm.

As the infant twitches in her sleep, nine thin wires taped to her face and bald head wiggle in all directions, giving her a baby Medusa look. Jenny’s mother opens her sleepy eyes in the dimly lit room and stares blankly into the tiny face only inches away. The matching wires on the mother’s head nod toward her baby as she unconsciously reaches out and pats Jenny reassuringly a few times. She adjusts the baby’s blanket, and they both drift back into a deeper level of sleep.

One room away James McKenna watches the needles on a 12-channel polygraph jump in tandem as Jenny and her mother experience this mutual arousal. An elfin grin spreads across his face. He’s recorded so many of these unconscious stirrings that they now seem to him to map out a nightlong dance.

McKenna, an anthropologist at Pomona College, has come to the nearby Sleep Disorders Laboratory at the University of California at Irvine to test a hypothesis: he believes that the Western practice of placing babies in their own beds at night is at odds with human nature--so odd that sudden infant death syndrome (SIDS), the mysterious killer of babies, can more easily come stalking. But he is just as interested in the vast majority of babies who don’t succumb to SIDS. Sleeping in isolation affects them too, he suspects, though more subtly than in the rare cases of SIDS. Jenny and her mother are providing the numbers to support what McKenna has been advocating for the past eight years: If you have a baby, sleep with it.

His idea developed from years of watching infant monkeys cling to their mothers day and night. He also knew that babies sleep with their parents in the vast majority of human cultures. Both facts suggested to McKenna that it’s inconsistent with our evolutionary roots to put babies in their own beds at night. What’s more, he points out, the current Western practice is only a century or two old, just a wink in human history. As an anthropologist with no formal medical training, however, McKenna hesitated to push for co-sleeping. Most pediatricians, after all, thought babies should sleep alone. Yet as he began to talk about his ideas, he found a receptive audience. His words, some parents told him, finally gave them permission to do what seemed to come naturally--sleep with their babies.

Many parents have fears about the safety of co-sleeping. They’ve been told that bed-sharing puts a squirming baby at risk of being suffocated by well-meaning but exhausted parents. This is probably no more than an old wives’ tale. As McKenna points out, most babies worldwide sleep with an adult without ill effects. Other parents feel that they need a break from the baby’s constant demands, or they crave time for intimacy. And current advice books uniformly reinforce the idea that sleep practices should accommodate parents, not babies.

Parenting advice in the 1990s, post-Dr. Spock, tends to be permissive. But in one area discipline survives: when, where, and how much babies should sleep. In The Well Baby Book, a popular guide, Mike and Nancy Samuels give parents hints to aid their quest to get tiny infants to sleep through the night. Don’t bring the baby into the parents’ bed and let it sleep there till morning, they say. It is more likely to be disturbed. Penelope Leach, in Babyhood, admits that babies sleep better when snuggled between adults. But Leach also writes that parents are often disturbed by the baby’s fidgeting, and many are uncomfortable with an infant in the marital bed. What’s worse, she and other authorities claim, co-sleeping establishes a dependency that will be difficult to break, making it hard for the older child to fall asleep when alone, although there is no evidence to support this.

McKenna believes the notion that solitary sleep is healthier for babies in the long run is based not on biology but on a recent adoption of urban-industrial values. Modern society requires good citizens-- independent people not making too many demands on others. In this scenario, autonomy must be fostered as soon as possible. We begin early, McKenna claims, by placing babies alone at night so that busy parents can get on with their lives. In our modern day, he says, the biological interests of the infants might not coincide with the best interests of the parents. But evolution never promised us a rose garden.

McKenna’s observation of mothers and infants began decades ago, with his training in primate behavior. As a junior at U.C. Berkeley in 1969, I took a course in primatology, he says. I learned that monkeys and apes need so much physical attention and contact. I remember thinking, when I have a baby, I’m going to give it as much affection as it can take. You cannot understand primates without coming to appreciate that very early physical contact is everything. It’s what we’re all about.

Now, at 43, he realizes that his later diversion to sleep research has even earlier origins. I grew up in a large family of six children. There weren’t enough beds in my house and we all shared beds. I slept at my brother Tommy’s feet for over a year! But it wasn’t until the birth of his son Jeffrey, in 1978, that McKenna put those research interests together.

I noticed that one way to get Jeff to sleep was to nap with him. I’d lie down with him and breathe as if I was asleep. He breathes in and out, in and out, pumping his chest up and down as if baby Jeff were still bundled on top of him. I became really skilled at getting him to sleep. I also found it totally amusing. But the scholar in McKenna was intrigued. I noticed he was so responsive to these breathing cues. And then I wondered why I was surprised. Here was a primate baby, undeveloped at birth, selected to be responsive to parental contact and care. The fact that he was responsive to my sounds and breathing patterns was everything the last ten years of anthropological research had told me he would be.

McKenna soon realized the implications of what he’d observed. In the United States one in every 500 babies is found dead during the first year of life, most often between two and four months of age. These babies usually show no previous signs of illness, and no known cause of death can be determined at autopsy--although recent research suggests that abnormalities in fetal development may predispose some infants to an early death. But it now occurred to McKenna that the absence of cues from co- sleeping parents might also play a role.

Since the medical community concentrates on physiological causes for SIDS, McKenna knew that any suggestion of a cultural influence would be considered radical. He knew he would need to explain how co-sleeping had evolved--how it contributed to a baby’s physical well-being. The difficulty is explaining to medically trained specialists what it means to apply evolutionary theory in the context of infancy and parenthood. That’s where I thought I could fill a role--making evolution alive and meaningful in the context of clinical research.
There were experiments McKenna could cite. As several psychologists showed in the 1960s, the infant’s physical dependence on its mother is a primate universal, and it involves more than simple providing. When infant macaque monkeys were separated from their mothers, even for a few hours, they experienced physiological effects such as changes in heart rate and body temperature, sleep disturbances, increases in cardiac arrhythmias, and signs of clinical depression. In short the animals’ immature nervous system just didn’t function as well. In the natural state, McKenna adds, monkey and ape babies always sleep with their mother, clinging to her belly until the infant initiates independence.

Human babies are even more dependent on adults. No other animal needs so much nurturing and takes so long to mature. The advantage to being so unformed at birth is the great capacity for learning and social interaction.

As with other primates, McKenna speculates, in humans the strong mother-infant bond was selected for because it helped babies get through their long formative period. But human babies are so helpless they can’t even cling to their mothers like monkeys. Instead, they are carried. In humans, McKenna says, infant sleep evolved against a background of being jerked up and down in the back of a sling. Even today you can see babies carried this way throughout Africa and Asia: mothers out hoeing in the garden, baby sleeping on their back. There is a physicality in the relationship, he says. We can’t go on assuming that there are no physiological consequences to sleeping alone.

McKenna suggests that all human babies benefit from hearing their parents breathe, feeling their parents’ touches, and just being close to adults. Although the long-term effects of solitary versus co-sleeping are unknown, McKenna suspects there’s a connection between nocturnal closeness and mental health later on, even into adulthood. A feeling of social- psychological connectedness allows infants to later become more independent from parents. It may also result in higher self-esteem and a good sense of empathy for others. These infants might also be able to better monitor nonverbal cues given by others.

More dramatically, McKenna believes, co-sleeping may be important in avoiding the particularly human problem of SIDS. Humans, he notes, are different from other primates in a way that makes us vulnerable: we rely so heavily on speech that we use voluntary, or controlled, breathing far more than any other mammal. We have to learn how to modulate our breathing to talk, though we never lose the ability to return to automatic pilot--the involuntary, reflexive breathing we use during sleep or reading.

Human babies begin switching back and forth from automatic to controlled breathing between two and four months of age. At this developmental stage, the infant neocortex, the higher brain, becomes functionally connected to the primitive brain stem. Behavior becomes less a series of reflex actions and more voluntary. Babies start to smile because they want to, and their vocalizations are no longer mere reactions to hunger or wet diapers. They begin to manipulate their breathing by changing airflow rates, air pressure, and lung volume. A cry will suddenly carry specific information to a carefully listening parent; it’s a form of speech-breathing that will later become talking. This is also a susceptible time for infants. Most do fine, but McKenna thinks some can’t manage the flip-flop between the two types of breathing. They stop, and succumb to SIDS.

To support his claims about the importance of co-sleeping, McKenna knew he would first need to show that babies are physically affected when they spend the night in contact with an adult. His co- sleeping hypothesis works only if infants sleep differently--presumably better--when tucked in with Mom.

One day in 1984 McKenna walked into the first open door in the pediatrics department at the University of California at Irvine and bent the ear of pediatrician Claibourne Dungy. Dungy quickly assembled what McKenna recalls as four people in white lab coats staring at me skeptically. One of them was Sarah Mosko, a clinical psychologist, a sleep expert, and most important for McKenna, a trained polysomnographer--a person who knows how to wire sleepers and interpret the squiggles on the polygraph. McKenna asked if she’d like to collaborate with him on his research. It didn’t take much for me to say yes, recalls Mosko, who now works as a sleep-disorders consultant in addition to her research with McKenna.

Together McKenna and Mosko have collected sleep data on eight mother and infant pairs at the Sleep Disorders Lab. In the first study, conducted from 1986 to 1987, five mother-infant pairs were tracked for one night. In the second study, finished last year, three mothers and infants spent the first two nights sleeping alone but in adjacent rooms, so that the mothers could get up and feed the babies. The third night each mother and baby slept in the same bed--an unusual event for two of the pairs.

Mothers and their infants report to the lab at 8:00 P.M. The sleep room, with a hospital bed and blackout curtains, seems to them an inviting haven. These are sleep-deprived new mothers, says Mosko. They usually say the time in the lab is the first reasonable night’s sleep they’ve had since the baby was born.

Before mother and baby settle down for the night, each has four wires taped to the head to record electroencephalographic, or brain wave, signals. Another two wires, placed close to each eye orbit, monitor eye movements, and three more wires on the chin measure muscle relaxation. Heartbeats are picked up by two wires on the chest. A thin wire placed beneath the nose monitors breathing by sensing the temperature of the passing air; exhaled air is warmer than inhaled air. (Breathing is also recorded as chest-wall movement.)

The data from all these sources help differentiate the five levels of sleep a person traverses during the night. Rapid eye movement, or REM, sleep is the most active--the eyes flicker, the face and limbs twitch as muscles tense and relax, brain waves come faster but with lower voltage, and breathing and heart rate become less regular. This is a dreaming state, although dreams sometimes occur in other stages. There are four non-REM levels; deep sleep occurs in levels three and four. Individuals vary in the amount of time they spend at each stage, and infants have fewer distinct levels--at three months they typically have three. In any case, several cycles through the various levels seem to be important for a satisfying snooze.

As McKenna and Mosko’s subjects sleep, impulses from the wires travel to the recording room, where inked needles leave tracks on long sheets of paper. Later Mosko gathers the pages and marks out sleep levels in 30-second intervals. She determines if each subject is sleeping at a certain level, awake, or experiencing transient arousal--moving into lighter levels of sleep but not to full wakefulness. McKenna, with his animal-behavior background, scores the videotape--baby lifts head, mother opens eyes, and so on. The two researchers eventually compare mothers and babies sleeping alone and together, interval by interval.

These data choreograph the nocturnal dance of mothers and babies- -the dance McKenna had predicted--but with mutual promptings and responses. It’s not that mothers regulate their babies’ breathing. The sleepers are, instead, physiologically entwined; the movements and breathing of each partner affect the other. When one arouses, the other often wakes up a bit, too.

McKenna proposes that transient arousals are especially important because they give babies practice in waking up. All babies experience apneas, or pauses in breathing, several times a night. If a pause becomes prolonged, a healthy baby will wake up to breathe. Many researchers believe that SIDS babies have some deficiency that inhibits their arousal. When they stop breathing, they’re less apt to wake up--and thus more likely to die. But if aroused more often by a parent, McKenna reasons, they may learn better how to do it on their own, and wake up one night when it really matters.

McKenna also suggests that co-sleeping helps a baby master breathing techniques. During sleep, just as during wakefulness, adults shift through periods of controlled or automatic breathing, switching between neocortical-driven breaths and brain stem-operated breaths. Babies undergo that flip-flop each time they wake up. When sleeping with Mom, a baby reacts to her movements and wakes up more times during the night--an average of 24 percent more, McKenna finds, than when sleeping alone--thus getting more practice in the repeated hop from one kind of breathing to the other. Sleep has evolved against these interruptions, says McKenna, and they may serve as practice for the baby when it has more serious internally based interruptions in breathing.

Such fitful sleep may, in fact, be the norm for adults as well. The mothers in McKenna’s experiments passed through transient arousals 60 percent more frequently when sleeping with their babies than when sleeping alone. We Westerners have the ‘die’ theory of sleep, McKenna says, laughing. You close your eyes, fall asleep, and basically die--you become totally unconscious until you wake up in the morning--and you hope for the best. If there’s anything in between, there’s something wrong with you. Other people in the world don’t sleep like that. The !Kung bushmen, for example, get up, tend the fire, talk, then go back to sleep. Western culture has streamlined what we think is normal. And if people can’t conform, there’s a disease out there for them--it’s called insomnia. A small group of sleep researchers have also admitted that humans are not monophasic sleepers--they are biphasic. The afternoon nap is biologically based.

His point is that cultures dictate norms unrelated to what might or might not be evolutionarily natural--that is, bred into human physiology. He feels that the extreme American emphasis on individualism, and the view that husband and wife have a relationship apart from the children, have reinforced notions that infants are born too dependent and should sleep by themselves as soon as possible. In contrast, Japanese infants normally sleep with their parents. This, too, is a culturally bound notion, but instead of opting for independence, the Japanese foster interdependence. Interestingly enough, the rate of SIDS is significantly lower in Japan than in the United States: less than one per 1,000 births.

Data for immigrant populations in the United States suggest that such cultural differences may indeed play a role. For example, Chinese immigrants in California have an incidence of SIDS 38 times higher than nonimmigrant Chinese in Hong Kong. Among other Asian-American populations the SIDS rates vary, but the rate increases the longer a group has lived in the United States. The Vietnamese, for example, arrived later than the Japanese, and their SIDS rate is lower. McKenna feels that the pattern may be explained by immigrants’ adopting the American style of placing babies in their own beds. His speculation cannot be confirmed, of course, until other possible influences--such as changes in feeding practices--are ruled out.

McKenna began giving talks about his ideas in the early eighties. Then, in 1986, he published a massive paper on his work, which attracted a lot of attention. So far the response from the medical community has not been as critical as McKenna first feared. Marian Willinger, who directs SIDS research at the National Institute of Child Health and Human Development, says, In general this is a new area for infant and child health--tying parenting styles with physiology--and therefore McKenna and Mosko’s basic research is important for all babies.

One medical researcher was deeply impressed. I think their work is terrific, says Jeffrey Laitman, an anatomist at New York’s Mount Sinai School of Medicine. His own research on the development of the throat and voice box in infants supports Mc-Kenna’s hunch that SIDS is linked to the evolution of speech making. In newborns, as in many animals, the larynx locks into the back of the nasal cavity, Laitman explains. This enables them to breathe and swallow at roughly the same time. But in humans the larynx begins to drop down into the throat in the first few months of life. No other mammal goes through such a tremendous metamorphosis, and there’s a great possibility of miscues--as well as a far greater ability to make the wide range of sounds used in talking.

But McKenna’s not just out to prevent SIDS; his approach has always been more anthropological than medical. His larger goal is to show that early sleeping practices are important to everyone’s health. This past January he and Mosko brought the first of 30 mother-baby pairs, including 15 co-sleepers, into the lab to investigate whether the sleep and breathing patterns of the co-sleeping babies are different from those of the babies who habitually sleep alone. McKenna expects to finish this study by the end of the year, but even then he’ll be a long way from proving that co- sleeping is best for everyone in the long run. His argument that it seems to work well in traditional cultures cuts two ways. After all, most American babies, with their background of solitary sleeping, also grow up apparently healthy.

For now, McKenna aims to prove that co-sleeping is natural and normal for the average baby, a reasonable option rather than a dangerous, misguided practice that should be discouraged, as stated in current advice books. Should a parent or parents feel good about co-sleeping, elect it as a favored strategy, and it is done responsibly, he writes, nothing could be better for their infant or child.

He is also philosophical about his potential role as a revolutionary in American parenting styles. There is nothing profound about what I am trying to document or argue for--it’s based on evolutionary history. It doesn’t take any genius to know there may be some naturalistic interactions between co-sleeping babies and mothers, or babies and caretakers. Like those who have discovered in the twentieth century that breast-feeding is good for babies, I spend all my time documenting the obvious.

Medical Research On...
Each year hundreds of papers are published on SIDS, pointing the finger at a host of possible culprits. Mothers who smoke during pregnancy, for example, have been told they’re upping their baby’s risk of SIDS about threefold. Babies may also be at higher risk if they are born prematurely or of low birth weight, as a sibling rather than a firstborn, or to a young mother. Babies who lie on their stomach have a higher risk; more babies die of SIDS in winter; elevated body temperature from a stuffy room or overdressing may be a factor.

Still, none of this explains the actual cause of sudden death. Risk factors are simply things that may make a baby more vulnerable, explains Marian Willinger, who directs SIDS research for the National Institute of Child Health and Human Development. Just because cigarette smoking is linked with an increased risk doesn’t mean that cigarette smoking causes SIDS. A lot of SIDS babies’ mothers don’t smoke. There’s something about the baby itself that predisposes it to SIDS.

Pinning down that something, however, has so far proved impossible. At this point, explains Willinger, SIDS is a diagnosis of exclusion. If you can’t find any other cause of death after a full postmortem, then it’s called SIDS--so by definition we’re starting without much to help us.

Nevertheless, some strides have been made. The most popular theory is that something is wrong with the way vulnerable babies arouse themselves from sleep--they’re supposed to wake up when they stop breathing for an unusual length of time, but they don’t. To investigate this idea, neurophysiologist Ron Harper and his colleagues at UCLA checked the records of nearly 7,000 babies whose heartbeats and breathing were recorded in a British study. Sixteen of those babies later died of SIDS; Harper found that they had gone through far fewer short respiratory pauses while sleeping than the ones who were still alive. Although the reason for this difference is not yet known, it is a true disparity.

Other researchers are looking at where respiration is controlled- -in the brain. The brains of all newborns are still developing; for instance, the neurons are not all covered by their protective sheaths of myelin. Early last year Hannah Kinney of Children’s Hospital in Boston and her colleagues showed that myelination in the brains of 61 infants who died of SIDS lagged significantly behind myelination in 89 children who died of other causes--though again this is so far just a clue.

Of course, a disease with such a nebulous definition can easily fool you. Researchers are fairly certain that 3 to 10 percent of SIDS cases are actually the result of inborn metabolic defects. And a study published last summer showed that a few babies diagnosed as succumbing to SIDS--fewer than 1 percent--might have suffocated on soft bedding such as beanbag cushions.

Yet researchers do feel that SIDS is a discrete entity with its own physiological mechanism, not just a conglomeration of other syndromes that simply need to be teased apart. The scientists really believe that after all is said and done there will be a core of babies with a certain characteristic abnormality that makes them vulnerable to sudden death, says Willinger. We won’t keep peeling away layers of onion until there is nothing left. -Lori Oliwenstein

Next Page
1 of 3
Comment on this article

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar

Log in to your account

Email address:
Remember me
Forgot your password?
No problem. Click here to have it emailed to you.

Not registered yet?

Register now for FREE. It takes only a few seconds to complete. Register now »